What a growth percentile means, and how to read the clinic's chart
You left the clinic with a number: the 23rd percentile. Or the 88th. The nurse said everything was fine, and on the way home you were already searching. Here is what that number really says, where it comes from, and what is worth doing with it.
Updated:
A percentile is a position, not a grade
A percentile compares your child with a hundred imaginary children of the same age and sex on the World Health Organization standards. The 23rd percentile for height means 23 of the hundred are shorter and 77 are taller. The 50th is exactly the middle: the median.
That is all the number says. It is not a grade, a ranking or a forecast. A child on the 10th percentile is not "doing worse" than one on the 90th, any more than an adult of 165 cm is doing worse than one of 185. They simply stand somewhere else in the normal range of human beings.
Where the curves come from
The curves your clinic uses, and the ones behind our calculator and app, are the WHO Child Growth Standards. For birth to five they were published in 2006 from a study that followed about 8,500 children in six countries (Brazil, Ghana, India, Norway, Oman and the United States) who were raised in good conditions and breastfed. The idea: describe how children grow when nothing is missing, rather than how they grow on average in one place.
For ages 5 to 19 the WHO 2007 reference takes over. The two sets meet at five, so one chart follows a child from birth to the end of school.
What a height curve looks like
In the chart below, boys from birth to five: the dark line is the 50th percentile, the median. The light band around it is the range from the 3rd to the 97th percentile, which holds 94 children in every 100. Notice how wide it is: at two years the gap between the 3rd and the 97th is about 11.5 cm, and all of it is the usual range. The girls' curve looks the same, a centimetre or two lower.
The shape tells a familiar story: babies grow about 25 cm in the first year, about 12 in the second, and then settle at 5 to 7 cm a year until puberty. That is why a month's difference in the measuring date matters a lot for a baby and almost nothing for an eight-year-old.
How to read the card
On the clinic card, as in the app, every measurement is one dot: age along the bottom, height up the side. One dot says little. Two start a line, and four or five show a direction, which is what the doctor or nurse is really looking at.
Three questions to ask the line: is it inside the band? Is it roughly parallel to the percentile lines, meaning the child keeps their place? And is there a sharp change, a fall or a rise that crosses lines? The third is the only one that usually leads to a closer look.
What counts as the usual range
Anywhere between the 3rd and the 97th percentile is the usual range. Below the 3rd or above the 97th is not a diagnosis either; it is a reason to measure again carefully and talk to the pediatrician. What draws professional attention is mostly a trend: a child who has slid over several measurements from the 75th to the 25th has crossed two major lines, and that is a common rule of thumb among clinicians for looking closer, even when the new number is itself "normal".
And a child who was born small and stays small? A child growing steadily along the 5th percentile is growing at their own pace. Much of it is inherited: short parents usually have short children. That is why the calculator and the app also offer an adult height estimate from the parents' heights, as a reference point rather than a prediction.
Measuring mistakes that move percentiles
One centimetre in a baby can move the percentile by ten points or more. The usual mistakes: measuring in shoes or thick socks; mixing lying-down length (up to two years) with standing height (after), since standing measures about 0.7 cm less, which is why the WHO curve deliberately "steps" at two; the time of day, because we are almost a centimetre taller in the morning than in the evening; and different people measuring with different technique.
So it is worth measuring in the same place, at the same time, without shoes, and writing down the date. One consistent measurement every three months is worth more than a weekly one under changing conditions.
Weight and BMI percentiles
Exactly the same idea with different tables. In children weight alone says little, so from age two BMI-for-age by sex is used, because the number itself changes naturally with age: a BMI of 16 is chubby at six and thin at sixteen. Here too: position and trend, not a grade.
What to do with the number
Write it down with the date, somewhere you will see the whole line and not only the last dot. Look at the direction. And if something does not add up, bring the card to the visit and ask: "Does this line look fine to you?" That is the right question, and it beats any search, this page included.
Check your child's percentile now
Date of birth, sex and height, and the free calculator places it on the WHO curve. In the browser, no sign-up, nothing is sent.
Open the percentile calculatorFrequently asked questions
The 50th is the middle, not the target. A child on the 20th who stays around the 20th is growing exactly as they should. There is no "good" percentile; there is a steady line.
Not from one measurement. Many babies "settle" onto their own curve in the first two years, and a fall or rise of one line is common. What justifies a word with the doctor is a fall that continues over several measurements, or crossing two major lines.
Roughly, and only from age two or three. Children tend to keep their percentile, but puberty shuffles everything for a few years. The usual estimate is from the parents' heights, and even that is a reference point with a range of several centimetres either way.
Probably a different table or different rounding. Some older cards use national or CDC curves, while the calculator uses WHO, as most clinics do today. A few percentile points between sources is normal; the trend over time within one source is what matters.
The guides here explain terms and are not medical advice. Percentiles are statistical context, not a diagnosis; for any question about your child's growth, talk to your pediatrician or clinic.